Abstract Immune responses to therapeutic proteins can pose enormous problems for the patient causing either adverse events or loss of efficacy. As therapeutic proteins represent one of the fastest-growing class of pharmaceuticals, it is vital to evaluate whether a protein is likely immunogenic early in the drug development stage and also to have methods available to re-design potential leads to eliminate immune response-causing epitopes. Equally important?yet demanding the opposite effect?is the design of broadly specific vaccines, where increased immunogenicity of subdominant conserved immunogens is desired. The adaptive immune response entails formation of a complex between a T-cell receptor (TCR) and a ?foreign? peptide bound to a major histocompatibility complex (MHC) molecule that is presented on an antigen presenting cell - a crucial step for the induction of high-titer IgG responses if CD4+ T helper cells become activated. De-immunization efforts of therapeutic proteins have mostly relied on experimental characterization of a large number of point mutants followed by laborious single peptide analysis assays. Data on of over 330,000 experimentally verified T-cell immuno-reactive peptides has been integration and used to predict new epitopes and to bias the computational re-design of a protein to eliminate T-cell epitopes. While this represents a promising strategy, improvements are still needed and no design approach to date is available to systematically incorporate optimized T-cell epitopes to increase immunogenicity. We propose to develop a novel computational design approach capable of reducing immunogenicity of therapeutic proteins or alternatively increasing immunogenicity of vaccine immunogens without disrupting proper folding or function. As proof of concept, we will demonstrate the ability of this approach to 1) reduce immunogenicity of a novel influenza protein therapeutic and 2) increase immunogenicity of influenza's hemagglutinin's stem-region. Finally, we will establish an in vitro high-throughput strategy to evaluate tiled peptide arrays derived from the re-designed proteins or viral surface proteins for the following characteristics: (a) immuno-reactivity, as defined by their ability to form a MHC-peptide complex and (b) ability to activate T-cells. Our approach will take advantage of oligonucleotide-chip technology and next-generation sequencing, enabling the screening of thousands of peptides in parallel. If successful, this strategy will provide a general computational protein design pipeline for reshaping the immunogenicity of proteins and immunogens as well as a two-level experimental verification platform for T-cell epitopes.